[mapserver-users] WCS time support bugs

[Armin Burger]

Alex

I had a look at the code and tried the suggested fix. I uploaded a patch 
at the wiki, see this link
http://trac.osgeo.org/mapserver/attachment/ticket/2487/mapwcs.c.patch

With these modifications the PostGIS problem disappeared and it was also 
possible to specify 'wcs_timeposition' as a range of dates instead of 
single dates for the full timeseries.

Armin


On 05/11/2008 17:46, Alexander Petkov wrote:
> On Mon, Nov 3, 2008 at 4:10 PM, Armin Burger <armin.burger at gmx.net> wrote:
>> Hi everybody
>>
>> I wanted to test the time support functionality of WCS but this seems not to
>> be working. There is an older bug report at
>>  http://trac.osgeo.org/mapserver/ticket/1856
>> but it seems this has not been fixed yet.
>>
>> I use a PostGIS table to store the tileindex for the time series and as
>> mentioned in the bug report there is still a wrong filter applied to the
>> tileindex. Did anybody already try if the fix suggested in the ticket will
>> work?
>>
>>
>> In addition, the time support of WCS does not have a time *range* similar to
>> WMS-T "ows_timeextent" like
>>  "ows_timeextent" "2006-01-01/2006-06-01"
>>
>> WCS requires to list every single date of the available coverages under
>> "wcs_timeposition" which is not very feasible for longer time series. The
>> bug report at Ticket #2487
>>   http://trac.osgeo.org/mapserver/ticket/2487
>>
>> lists this problem as fixed, but when I use
>>   "wcs_timeposition" "2006-01-01/2006-06-01"
>>
>> and request a coverage with a parameter '&TIME=2006-05-11', with MapServer
>> v5.2 I still get the exception
>>  ...
>>  WCS server error. The coverage does not have a time position of
>>  2006-05-11.
>>  ...
>> Any idea if the definition has to be set in another way?
>>
>> Best regards
>> Armin
> 
> 
> 
> Hello Armin:
> 
> I experienced the same problems some time ago. Here are the relevant
> posts to the ML that I wrote:
> 
> http://www.nabble.com/wcs_timeposition%3A-explicit-listing-or-start-end-range--td19001008.html
> 
> And the problem with using PostGIS connector:
> http://www.nabble.com/GetCoverage-request-and-time-idexed-coverage-using-postgis-td18743297.html
> 
> I've been told to file a bug, (which I have yet to do), I hope to get
> some time soon to look into the code myself and see what's going on...
> 
> Alex
>